March 31, 2010
Considering putting a hyperlink on your website or blog? The case Crookes v Newton, 2009 BCCA 392 [Crookes], may give you pause to reconsider. Although the British Columbia Court of Appeal ("BCCA") held in Crookes that a website owner will not be liable for hyperlinking to defamatory sites, the decision left a number of unanswered questions that may come into play when the Supreme Court of Canada ("SCC") decides whether to grant Wayne Crookes leave to appeal on April 1st, 2010.
Background
Mr. Newton is the owner of the website P2Pnet, whose content is being challenged as defamatory. In 2002, P2Pnet went online, claiming to be the first Internet webpage to give daily updates, news, stories and commentary without being “spun, filtered, and pre-digested by vested corporate interests”. His website contains commentary on various issues, including issues involving free speech and the Internet.
In May of 2006, Mr. Crookes began a defamation action against an author for publishing a series of online articles that he calls offensive. Mr. Newton subsequently published a piece on P2Pnet, detailing Mr. Crookes' action and its implication for those who operate Internet forums. He did not reproduce the content of the offending article or comment on it in any way, but merely linked to it on his website. Mr. Crookes and his lawyers contacted Mr. Newton to remove the link from his site, but Mr. Newton refused to do so.
Basic Principles
The BCCA found for Mr. Newton, holding that he was not liable for linking to the allegedly defamatory articles. The court considered three main issues: (i) whether or not hyperlinked articles attract a presumption of publication, (ii) whether hyperlinking is a form of publication, and (iii) whether the volume of hits of a website can lead to an inference of publication.
(i) Mere Hyperlinking Does Not Give Rise to a Presumption of Publication
If the act of hyperlinking gave rise to an automatic presumption of publication, it would would shift the onus onto the defendant to establish that the hyperlink did not, in fact, constitute a publication. But both the majority and dissent concluded that a hyperlink does not attract an automatic presumption of publication. The fact that a website owner put up a hyperlink does not automatically make the link a publication.
The BCCA then considered the question of whether hyperlinking in this particular situation constituted a republication of the allegedly defamatory articles. The court relied on a previous decision, Carter v BC Federation of Foster Parents Assn, 2005 BCCA 398, where it was held that the publication of a website address does not result in republication of the material on the website. Following this precedent, the court in Crookes held that there is “no substantial difference between providing a web address and a mere hyperlink.”
(ii) Circumstances Where Hyperlinking Amounts to Publication
Where the majority and dissent differed was whether the hyperlinking in this case actually amounted to a publication. The dissent set out various factors to consider when determining if a hyperlink was a publication, such as
“the prominence of the hyperlink, any words of invitation or recommendation to the reader associated with the hyperlink, the nature of the materials which it is suggested may be found at the hyperlink (for example, if the hyperlink obviously refers to a scandalous, or obscene publication), the apparent significance of the hyperlink in relation to the article as a whole, and a host of other factors dependant on the facts of a particular case.”
Words of encouragement or invitation that accompany the hyperlink may lead to a finding that the hyperlink was a publication. As the Court stated:
“Whether the hyperlink is a web address, as is often the case, or a more specific reference, both require a decision on the part of the reader to access another website, and both require the reader to take a distinct action, in the one case typing in a web address and in the other case clicking on the hyperlink. In other words, there is a barrier between the accessed article and the hyperlinked site that must be bridged, not by the publisher, but by the reader. The essence of following a hyperlink is to leave the website one was at to enter a different and independent website.”
In this case, Prowse J.A., in dissent, held that the Mr. Newton’s piece on P2Pnet, as a whole, contained words of encouragement and invitation that led to the conclusion that the hyperlink was a publication.
The majority disagreed with that conclusion, finding that Mr. Newton did not comment on or reproduce any of the allegedly defamatory material and did not encourage the reader to follow the link to the sites. His articles could be considered to be statements of approbation, or adoption, but were more similar to a footnote for a reader or a card index in a library. The separation between the reader and the defamatory material led the courts to conclude that the hyperlink was not a publication.
(iii) Volume of Hits to a Website Is Insufficient Basis to Infer That User Followed the Hyperlink
The majority also concluded that the number of hits of Mr. Newton’s article was insufficient basis to infer that the user followed the hyperlink. It stated that while there may be cases in which web traffic could support such an inference, this was not one of those cases. The dissent, however, found that on the facts of this case, one could infer publication based on volume of hits.
Implications
This case carries great significance for current and future website owners. Prowse J.A.'s dissent expressed a balance between the competing values of freedom of speech and protection of reputation. The relative weight given to both concepts in the defamation context should be “value laden” and “policy-directed.”
Many critics have noted that although this may be a win for Mr. Newton, the impact to bloggers and website owners is uncertain because while the majority didn't find any liability, it still suggested that hyperlinking to third party content could, under certain circumstances, leave a website owner vulnerable to a defamation suit. The decision leaves many questions unanswered about the circumstances in which a creator of a hyperlink will be found liable for defamatory materials found at a linked site. The decision is also unclear as to what are the implications are for website owners who refuse to remove the hyperlink when asked to do so.
The decision in Crookes presents the SCC with an opportunity to clarify the nature and scope of liability for hyperlinking to third party content. If, however, the Supreme Court decides to deny leave to appeal, the need to reform libel law still stands, especially with regard to Internet communication. Perhaps this area of the law could be clarified through legislation, following in the footsteps of US law.
6 Comments
I agree with you regarding the 'victory' in the BCCA Judgement. I was left with a definite sense that hyperlinking to third party content could, under certain circumstances, leave a website owner vulnerable to a defamation suit. As you rightfully state, this case gives the opportunity for the Supreme Court to clarify the nature and scope of liability for hyperlinking to third party content.
As you say it doesn't' obviate the need for libel reform but the uncertainty needs to be removed for bloggers and others who use the internet as a publishing platform.
To me this is one of the most significant cases regarding freedom of speech.
We have seem some hyperlinking copyright cases (eg Cooper case in Australia) but here we are talking about libel by hyperlink.
It is scary stuff particularly in the context of the facts of the case and the precedents that existed in the offline world. I would have thought that the Court would have stated emphatically that there could be no publication by inference in this particular instance.
This is an extremely poor summary of the BCCA ruling.
Presumption of publication was not denied only for hyperlinks, but for *ALL MATERIAL THAT APPEARS ON THE INTERNET* at least without some additional proof that someone in BC saw it. That is quite clear from the following phrases from the majority:
" In coming to the conclusion that there was no presumption of publication on these facts, the trial judge relied on Crookes v. Holloway, 2007 BCSC 1325 (CanLII), 2007 BCSC 1325, 75 B.C.L.R. (4th) 316, aff’d in Crookes v. Yahoo! Inc., 2008 BCCA 165 (CanLII), 2008 BCCA 165 (sub nom. Crookes v. Holloway), 77 B.C.L.R. (4th) 201 [Holloway]. "
None of those cases involved hyperlink indirection.
"With respect to Mr. Crookes’ argument that there should be a presumption of publication, Mr. Justice Lowry, speaking for the Court, stated, at para. 6 of the decision: [6] ...the statements that are the subject of Mr. Crookes’ action were posted on a website with restricted access that was not available to the public. The basis for any presumption that might be said to have been recognized in Wiebe does not exist here. I do not consider the mere fact a statement was posted on a website with the kind of restricted access there was in this case supports the presumption it was read by anyone in British Columbia. [Emphasis added.]""
So the decision applies not only to hyperlinks that may not have been traversed but to all situations where there is restricted access.
Furthermore the majority makes clear that situations where there are few or no proven viewers are also denied presumption:
"[66] Since **there is no presumption of publication with respect to Mr. Newton’s article**, including the articles accessible via the hyperlinks,"
[Notice that Newton's article has no presumption and that the articles "accessible" *ALSO* have no presumption]
"...the onus was on Mr. Crookes to prove, on a balance of probabilities, that the hyperlinks contained in Mr. Newton’s article (or at least one of them) were accessed by at least one person who read the article or articles. Such proof does not require direct evidence from a third person who used the hyperlinks to access the impugned articles. [While] It is not necessary for the plaintiff in every case to prove directly that the words complained of were brought to the actual knowledge of some third person. If he proves facts from which it can reasonably be inferred that the words were brought to the knowledge of some third person, he will establish a prima facie case. "
The majority concluded that the evidence of 1788 hits to Newton's page was insufficient to establish that anyone had read them, which is correct as a great many of the hits are typically from software robots. Prowse disagreed and said:
"[70]... it seems unlikely that of the 1,788 who chose to access Mr. Newton’s article on free speech, not one would have chosen to access and read the impugned articles contained at one or more of the hyperlinks. (In coming to this conclusion, I accept that some of the “viewers” of the p2pnet site may have been software robots.)
[71] While the evidence of invitation to visit the hyperlinks was more subtle than a statement in the original article that “the truth about Crookes can be found here”, I find that the number of “views” of the original site, the fact that the article deals with free speech and defamation, and the reference to lawsuits involving Mr. Crookes, would have served as words of encouragement, or an invitation, to a person viewing Mr. Newton’s article to look further. In other words, I do not accept that the hyperlinks, when viewed in the context of the Newton article as a whole, operated as a “mere” bibliographical” footnote.
[72] In summary on this point, I find that the trial judge erred in failing to draw an inference on all of the evidence that Mr. Crookes had established publication of the hyperlinked articles to at least one third party. "
It's the question of evidence that is required for *ANY* Internet documents to be considered publication, not merely those that are behind an inter-domain hyperlink, that's at issue. Notice also this reference to English vs. American law:
"[27] In Carter, this Court noted that the relative weight to be given to the concepts of freedom of expression and protection of individual reputation in the defamation context is one which is value laden and policy-directed. In preferring English and Australian authorities over the “single publication rule” adopted in the United States Mr. Justice Hall, speaking for the Court in Carter stated, at para. 20:
[20] Ultimately, the question of whether the Commonwealth or American approach is preferable has a considerable element of policy inherent in it."
The English approach is to set a high standard of publication but allow actions on a wider range of statements, while the American approach is to set a low standard of publication but to restrict actions to those which meet a high standard of malice, proven damages, specificity to a particular plaintiff (not a group) and lack of public issue relevance.
The judgement also makes reference to the ICCPR, Charter and other binding law.
While it's less than explicit in the judgement, there is also direct reference to Mr. Crookes' proliferation of actions and the issue of deliberate suppression of those critical of himself - the fact that Newton was sued effectively *FOR* suggesting that his readers should support Mr. Piling in a related case.
The idea that this case is only about material behind inter-domain hyperlinks from one publisher to another is simply wrong. The appeal to the Supreme Court was focused on that, but it's not at all clear why the SCC chose to hear it, and they may comment on any and all of the above issues, including Prowse's threshold.
Three other issues of note that you neglect and should rewrite the article to fix:
1. The actual difference between the majority (Bauman, Saunders) and minority (Prowse) was whether Newton's total refusal to modify material or respond to any of Crookes' concerns, and Newton's sole control of the material making it impossible for Crookes to get any other remedy, was sufficient to cause the trial judge to look at the other circumstantial evidence for publication, i.e. other than affidavits from persons who saw it in BC.
The minority (Prowse only) thought this factor decisive. The majority (Bauman, Saunders) agreed it was very important and should be considered but found it impossible to reconcile the lack of presumption of publication with a requirement to respond to complaints - if the material is not published and not proven to be, why would the non-publisher have to respond to complaints about the non-publication? He is within his rights to ignore such complaints.
2. Crookes actual position is not just that one link leads to liability but that (as http://pacelegal.com.au/tag/crookes-v-newton/ correctly reports in a much briefer and better report)
"Mr Crookes attempted to persuade the nine justices that linking to a link that links to a link ad infinitum is publication, quite a startling proposition (at least in my mind)"
2a. This position actually assumes a uniform semantics of transitive legally binding approval for every link on the Internet, which is more "startling" than it seems, as even links that imply disavowal (such as links to archived/old versions that have been explicitly withdrawn or corrected, kept for records purposes only) might be interpreted as republication. So much for Wikipedia's page version history! To let arbitrary parties read these archives becomes a totally unacceptable legal risk, as they are necessarily full of incorrect statements and unjustifiable claims.
Your coverage should focus on the dangers of the SCC trying to invent some kind of semantic standards for the web, when decades of academic attempts to do so have failed even to achieve a standard means of signalling what those semantics are. (the "REL=" and "REV=" tags in links are about as close as anyone ever got, other ontology standards simply do not apply to every static HTML web page).
3. Furthermore Crookes continues to assert that he can presume publication on the Internet under a US-style "single publication rule", despite constant denials of this by the appeals court. When directly asked by Justice Leask of the Supreme Court of BC how the pending Newton ruling would affect another case (West Coast Title Search and Wayne Crookes v. Pilling et al), Kasting's response was that there was nothing pending that would. Meaning that even if the SCC rules for Crookes in Newton, the case still cannot continue for lack of this proof, unless the SCC agrees with Prowse and establishes some alternative route for proving publication on the Internet, based perhaps on calculations of hits and probabilities of "click-throughs".
I will review Crookes' factum again to be sure but unless an officer of the court of BC is telling the Supreme Court of Canada one thing and the Supreme Court of BC another, then the question of presumption of publication is no longer at issue. Thus (as Pace Legal assumes) only when the publication (receipt of the material in the jurisdiction) is proven by some other evidence can each step of the infinite transitive chain of liability be followed.
This may not be easy to understand, so let me use an example:
Imagine a page containing a link to a cited source that contains a link to a news report that contains a link to archives of older versions of that same news article.
Now someone claims that an older version of the article contained defamation. The operator of the web site with both the news reports and older versions says it does not, but in any case was withdrawn and rewritten already. The complainant objects to the mere existence of the archive (exactly the situation in Pilling et al). Despite the corrected version being the only one endorsed for six months, sues not only the archive and news maintainer but the sources that link to it and those that link to *that*. (In Newton, usgovernetics.com and Pilling's report of the case are two "cited sources", Pilling's openpolitics.ca has the "news report" (deep within it) and the archives)
To prove liability even in a regime where liability is transitive, given the lack of presumption of publication on the Internet, plaintiff must at least prove that the inter-domain links were used to transit from the initial page to the cited source(s) to the news report and thence the archive. This is apparently not at issue at SCC.
In the (theoretical) circumstances where such proof exists, then the court can examine the semantics of any particular link. If all of them imply approval of the material (which the archive clearly does not, but the court may make an error) impugned, then, there is a need to hear other defenses of that material in the common law (reverse onus). In Quebec Civil Code fault must be shown and the onus is on the plaintiff.
So the SLAPP, ICCPR, Charter and uniform justice issues are important. Both Newton and another case, Editions Ecosociete, may contain some guidance on that problem.
In my opinion (only) plaintiffs are seeking a route to apply endless process to deep pocket defendants who will pay them to go away. Google being chief among them.
Thank you Craig for your helpful and thorough legal commentary on this defamation case.
Who has irrevocable rights to material once it has been published on the internet?
What level of technolaw security is warranted to protect original material from being hyperlinked to unauthorized third parties? Would our copyright laws apply in this case and would they be adequate protection against unauthorized use?
As you have stated, the US has tried to deal with the issue of liability under the Communication Decency Act, CDA, 47, USC, Section 230 to limit "a route to apply endless process to deep pocket defendants who will pay them to go away". Does this Act apply to hyperlinking too?
It will be interesting to see how the courts deal with this tangled web of transitive liability.
Sona, I hope you will continue to review and update your article about this challenging case.
It does not seem to me that any transitive liability rules applicable to libel/defamation can or should apply to copyright. For one thing explicit disavowal/disapproval that has no effect on the accessibility of the material, for instance which does not move it back behind a list of "archived" materials or otherwise make it harder to access, would not be as much of an issue in copyright as in defamation. While one can say "look at this terrible thing done over here, a copy of __LINK_TO_MAJOR_MOTION_PICTURE_BOOTLEG__ free for all to download oh my how awful, what lawbreakers!" and the anchor text surrounding it would be ignored, that is simply not the case for libel/defamation liability. If I say "look at how Stephen Harper has been accused of bribery and perjury and stealing an election, with all __THIS_EVIDENCE__ linked __FROM_THIS_COMMENT__, oh my how terrible", I might actually mean to disavow it, or endorse it, but explicitly, I have said I disavowed it, and have to be taken at face value. Else no reportage on anything controversial can be safe.
Too much has been made of this transitive liability issue as it may apply to copyright. Who cares? It's a marginal question. I care whether civil suits can be used against Google to expose activists' email to brutal Arab or Asian dictatorships. I care whether public figures engaged in crime can be accused of it, with the evidence, in open media, or whether they get to suppress the accusation with their political ties (as without doubt has happened in the Cadman/Harper case).
In these Crookes cases, we should ask whether Canada will endorse US-style strict pre-emptive "Sullivan" criteria (specificity of comment to a particular person not a group of jointly defamed persons, proof of actual damages, proof of malice, lack of public issue relevance) or endorse Quebec- or California- style anti-SLAPP provisions as a consequence of the protections for free speech in international laws Canada is party to. If it does neither and merely meekly moves the threshold of publication around a bit, or (worst) endorses this foolish view of Prowse's that when a comment is both priveleged and unpublished, that intransigence in editing it is to be punished, then we can simply look forward to more absurd cases like this at SCC.
Any further attempt to "refine" English common law of libel rather than sharply restrict it's application, is in my view a logical error. Canada would be alone in such "refining" among all English speaking countries other than England. I'd say there is no logical stopping point in law short of the Sullivan or SLAPP solution, the English system (it does not apply in Scotland) is a mindless unprincipled joke only slightly moderated by assuming a higher threshold of proof of publication.
If the SCC endorses the plaintiff-friendly English model rather than Quebec's anti-SLAPP or the ICCPR pre-emptive "arbitrary and unlawful" criteria, it makes it very difficult to justify running any political comment blog or website from anywhere in so-called "English Canada". It would be wiser to base them in Iceland, or at least California or Quebec, and make all postings from jurisdictions that enforce reciprocal judgments from Canadian civil courts, invincibly anonymous. If Crookes gets anywhere, detailed instructions for that will emerge instantly... and extinct the jurisdiction of "our" courts over literally any form of communication. Free speech on political issues outweighs the right to be free from the negative effects of literally any other kind of free speech, so if it comes down to that choice, the world will accept more untraceable death threats and child pornography in order to ensure more Arab Springs in the future. History says the lack of free speech kills far more people than harmful speech does. The SCC may force such blunt calculus simply by endorsing Prowse's self-contradicting position, or requiring one affidavit from just one "offended" viewer to prove publication, both open to process abuses.
It would be better to have the courts actually uphold the priveleged position of free political speech (note that in Prowse's comments she is inexplicably unaware of the irony of arguing that people trying to cooperate to protect free political speech as their objective should be at more rather than less civil risk, this is pretty amazing) so that they could retain jurisdiction over all speech and not encourage anonymizing absolutely everything, but the SCC made a grave error in 1995 in Hill v. Scientology and has compounded it since. It's long past time to reject Peter Cory's assessment that somehow there was "no evidence" of libel chill effects in Canada. Certainly there is *now*: Chuck Cadman, anyone? Editions Ecosociete?
Such an SCC error would also constitute a good first step towards the final political separation of Quebec, as a national border would make it much easier for Quebec to deny SLAPPs against media organizations based there, even if written in English for Canadians. SLAPP filers like Wayne Crookes or Stephen Harper would be perceived, correctly, as doing extraordinary damage to uniform civil justice and freedom of political comment in Canada, if more suits like that against Editions Ecosociete get filed. Sanctioning such people is a very good reason for separation from the entity ("Canada") that sanctions them, for those who like international law.
Harper gets to appoint up to five Justices in his first majority. The only time to get a rational position on political free speech into Canadian law would thus be *now*. Future justices we can expect to sympathize with SLAPP filers and political libel chill.
Hello I'm not sure whether it's me or simply yuor web blog
but it's loading slow for me, I had to spend just like a couple
of minutes to successfully load up still , facebook operates
perfectly . Well, Thank you for placing a remarkably brilliant
articles. In my opinion it really has already been honestly helpful
to individual who seem to visit here. I personally ought to state that you really have done brilliant work with this and additionally wish
to find even more awesome things from you. To obtain more knowledge
through content which you write-up, I've saved to bookmarks the
site.